PROJECT SUMMARY Proteins are macromolecules that carry out innumerable biological functions ranging from intracellular signaling to immune response. The specific function of a protein is determined by its three-dimensional structure. Therefore, an essential step towards understanding protein function is elucidating protein structure. However, the premier method for structure determination, X-ray crystallography, can be limited by many potential factors including the quality of the protein crystal. The proposed research will investigate the effects of protein? protein interactions in the crystal lattice on crystal quality and resolution and develop a design approach to stabilize lattice interactions, which should result in higher-quality crystals. First, the molecular determinants of high-quality protein crystal structures will be identified by (1) curating a set of representative structures from the Protein Data Bank, (2) extracting relevant features, such as interaction energy, buried solvent accessible surface area, interfacial packing quality, and residue usage, and (3) analyzing the features? relationship to crystal resolution (a proxy for quality). Based on these findings, a design strategy and score function for ranking designs within the Rosetta framework will be developed. Second, the design strategy will be applied to SNase, a model protein that is easy to purify and well- behaved in crystallization and diffraction experiments. The designed proteins will be crystallized and the crystal structures will be solved, testing for improved resolution. Analysis of the resultant crystal structures will drive development of the design strategy. Third, the design strategy will be applied on a bacterial gyrase, an antibiotic-target protein for which there are several drug-bound structures at low resolution, lacking sufficient detail to reveal key antibiotic? gyrase interactions. Preliminary data suggests that the designed gyrase mutants will yield high-resolution structures, permitting a better understanding of antibiotic?gyrase interactions, with implications for drug design. Should the method be successful, it will be immediately applicable to ~26,000 structures in the Protein Data Bank, and countless structures that have not been published due to lack of resolution. Re-engineered, high-resolution structures of these proteins could yield structural data on molecular interactions pertinent to disease, drug development, and basic understanding of protein function.